chapter one metal-directed macrocyclization reactions ... · work involving condensations using...

46
1 Chapter One Metal-Directed Macrocyclization Reactions Involving Formaldehyde, Amines and Mono- or Bi- Functional Methylene Compounds 1. Introduction The synthesis of large saturated organic rings incorporating a number of heteroatoms presents a challenge for synthetic chemists which has been approached in a number of ways. A number of routes to macrocyclic molecules have been developed employing standard organic chemistry approaches 1 . In all reactions, an important step must be the cyclization process, but this is not entropically favoured when forming large rings, and consequently reactions may occur in low yield or, in order to promote higher yields, must be accomplished under high dilution conditions where the cyclization is more favoured. Nevertheless, some macrocycles can be prepared conveniently in reasonable yields by direct organic routes. However, it has been recognised for several decades now that it is possible to make use of the 'organising' capacity of metal ions to promote reactions which lead to macrocyclization 2 . One consequence of binding some reaction components to a metal ion is that co-ordination reduces the large-ring reactions occurring in 'pure' organic routes to small-ring reactions where the building of small (typically six-membered) rings incorporating the metal ion is

Upload: others

Post on 16-Feb-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

1

Chapter One

Metal-Directed Macrocyclization Reactions

Involving Formaldehyde, Amines and Mono- or Bi-

Functional Methylene Compounds

1. Introduction

The synthesis of large saturated organic rings incorporating a number of

heteroatoms presents a challenge for synthetic chemists which has been

approached in a number of ways. A number of routes to macrocyclic molecules

have been developed employing standard organic chemistry approaches1. In all

reactions, an important step must be the cyclization process, but this is not

entropically favoured when forming large rings, and consequently reactions

may occur in low yield or, in order to promote higher yields, must be

accomplished under high dilution conditions where the cyclization is more

favoured. Nevertheless, some macrocycles can be prepared conveniently in

reasonable yields by direct organic routes.

However, it has been recognised for several decades now that it is possible to

make use of the 'organising' capacity of metal ions to promote reactions which

lead to macrocyclization2. One consequence of binding some reaction

components to a metal ion is that co-ordination reduces the large-ring

reactions occurring in 'pure' organic routes to small-ring reactions where the

building of small (typically six-membered) rings incorporating the metal ion is

2

then important. Formation of the small chelate rings is favoured entropically,

in the same way that small-ring formation reactions in organic chemistry are

also facile. There are a number of ways in which metal-directed reactions can

occur, but dominantly they involve the reaction of two heteroatoms bound in

adjacent sites in the co-ordination sphere of a metal ion with other reagents

which lead to formation of a new chain linking the originally separate

heteroatoms. Where a sufficient number of these reactions occur around a

metal ion, a macrocyclic molecule ligated to the metal ion can result;

alternatively, a larger polydentate but not cyclic molecule can result if reaction

is limited in some way.

Metal-directed reactions leading to macrocyclic molecules are reasonably

diverse, and it is the intention of this review to devote itself largely to

reactions which produce saturated macrocyclic ligands and related polydentate

molecules and involve formaldehyde as a reagent. Moreover, incorporation of

additional reagents in the reaction generates alternate products. Where

monofunctional or bifunctional methylene compounds of sufficient acidity are

involved in addition to formaldehyde, these carbon acids may be incorporated

into new rings along with formaldehyde, generating macromonocycles which

contain additional pendant groups which may carry potential donor groups.

These pendant donor groups increase the capacity of ligands to saturate the

coordination sphere of metal ions, and add an additional 'dimensionality' to the

macrocycle as a result of their spatial orientation relative to the donors in the

macrocycle.

3

1.2 Reactions Involving Formaldehyde Alone

Coordinated primary amines present the most appropriate sites for reaction

with formaldehyde via reaction as in eq. (1).

R-NH2 + CH2O R-N=CH2 + H2O R-NH-CH2-OH (1)

where the imine an the aminol may exist in equilibrium. Although coordinated

imines are more stable than free imines, and have been isolated and

characterised in some circumstances3, they remain susceptible to further

nucleophillic attack. Where no other added nucleophile is available, further

reaction with accessible components of the precursor complex, such as an

additional primary amine, may occur. Reactions between coordinated

molecules and formaldehyde alone have received only modest attention

compared to some other metal directed reactions. However, they represent a

useful starting point for this review, and are addressed below.

One of the earliest reactions involving a polyamine coordinated to a metal ion

and formaldehyde was reported by Alcock et al.4. The cyclization involved the

reaction of a tetrahyrdrofuran solution of L1 with formaldehyde and nickel(II)

salts in air to give the macrocyclic nickel(II) complex of L2. Condensation of

one formaldehyde molecule with one primary amine has been followed by

nuclelophilic attack on the resultant imine (or aminol) by the adjacent primary

amine, the intramolecular reaction generating a new chelate ring and

4

completing cyclization. A similar condensation was reported for co-ordinated

dihydrazines by Peng et al.5

Steric strain in complexes of ligands such as L2 is not marked, as six-

membered chelate rings form. However, it has been observed that adjacent

coordinated amines can be linked directly in the same manner. Geue et al.6

have shown that cis-disposed amines may react with formaldehyde to form

diaminmethylene (>N-CH2-N<) links in several cobalt(III) polyamine

compounds resulting in 4-membered chelate rings. Despite the strain present

in the small chelate rings, formation reactions are surprisingly facile.

A reaction which produced a macrocyclic ether complex with a coordinated

oxygen around a copper(II) template has been reported7 . This was achieved by

the addition of formaldehyde to a reaction mixture of 4,7-diazadecane-1,10-

diamine, 2-aminoethanol and copper(II). Four intramolecular condensations

occur leading to four new links identified in bold in L3. Apart from >N-CH2-N<

5

formation, attack of the ethanolamine oxygen on an aminol has produced an

ether via a new -O-CH2-N< bridge.

A range of interesting reinforced hexaaza macrocycles with two additional

ethylene bridges joining adjacent nitrogens (L4, L5, L6, L7) were synthesised

by Suh and co-workers8,9,10. These were produced by reacting

polyaminoalkanes (see eua. (2)-(5) below) with formaldehyde, using nickel(II)

as a template, with the variation that both ethane-1,2-diamine and a

tetramine were present in the later two synthesis. It is notable that all

reactions produce a 14-membered basic macrocycle framework, a ring size

which is generally the most thermodynamically stable in complexes with first

row transition metal ions. Again, in L4-L7, new links completing the

macrocycles are shown in bold. The similarity to the synthesis forming L3,

where again a 14-membered basic ring is formed, are obvious.

6

Ni + [Ni(4)]2 2 2

Ni + [Ni(5)]

Ni + [Ni(6)]

Ni + [Ni(7)]

2+

2+

2+

2+

2+

2+

2+

2+

H N NH NH CH O+

2 2H N NH NH + 2CH O

2H N NH NH 2

NH +2CH O 2

H N2

NH+

2H N NH NH

2NH +

2CH O

(2)

(3)

(4)

(5)2H N2

NH+

Alternative condensation reactions involving two formaldehyde molecules in

concert are known. Also in the early seventies Brush et al.11 found that bis((S)-

serinato)copper(II) L8 reacted with excess formaldehyde at pH 7-9 to form the

complex L9. In the absence of the metal ion no reaction was found to take

place. Presumably, acidity of the amino acid methylene in the coordinated

complex is sufficiently high to act as a nucleophile along with the alcohol arm.

Three formaldehyde molecules are incorporated in each amino acid unit, as

indicated in bold in the drawing of L9.

7

The condensation of formaldehyde with glycinatobis(ethylenediamine)-

cobalt(III) to yield [(-hydroxymethylserine)bis(ethylenediamine) cobalt(III)]2+

L10 and the macrocyclic species [(-hydroxymethylserine) (1,4,8,11-tetraaza-

6,13-dioxacyclotetradecane) cobalt(III)]2+ L11 was the first reported

condensation of this type around a octahedral metal ion12. In the latter

compound, apart from attack of a formaldehyde on the glycine, new chelate

rings containing an ether oxygen are formed by attack of the oxygen of one

aminol on an adjacent imine carbon, both arising initially from primary amine

and formaldehyde condensation.

8

A range of macrocyclic oxatetraazacycloalkanes copper(II) complexes (L12,

L13, L14, L15) have been produced in which the oxygen, while still a member

of the macrocyclic ring, was not coordinated to the metal ion13. They were

produced by the condensation of two formaldehyde molecules with two cis-

disposed amines in the copper(II) complexes of the precursor

polyaminoalkanes, as in L11 above. New links formed are shown in bold in L12

L15. The procedure was only found to be effacious when the final product

was essentially a 14- or 15-membered macrocycle; attempts to perform the

equivalent chemistry to produce 13- or 16-membered macrocycle were

unsuccessful.

1.3 Reactions Involving Formaldehyde and Ammonia

A coordinated imine is highly activated towards nucleophilic attack, and we

have seen in the previous section how an adjacent coordinated amine can act

as an efficient nucleophile. It is hardly surprising, therefore, to find ammonia

itself, as well as uncoordinated primary amines, can participate in reactions in

concert with formaldehyde and coordinated amines. In effect, ammonia

9

functions as an acid in these reactions, releasing a proton and permitting the

amide ion to attack the imine, as in eq. (6), where coordination of the imine

nitrogen is implied.

R-N=CH[Co(L17)]3+ [Co(L17)]3+ + -NH2 + H+ R-NH-CH2-NH2 (6)

The primary amine terminating the new chain still carries two protons, and in

principle can undergo two additional similar reactions until converted to a

tertiary nitrogen. This capacity to act as a nucleophile in three successive

reactions was first utilised in reactions to form macrobicyclic polyamines,

members of the cryptate family of encapsulating molecules.

Ligands that effectively encapsulate metal ions have received a great deal of

attention since the early seventies and an important class of these ligand are

the cryptates. Sargeson and co-workers have used formaldehyde in concert

with other reagents as and important synthetic reagent to produce a range of

cryptates of varying structures, physical and chemical properties, employing

inert complexes as precursors.

One of the earlier of these ligands produced was the macrobicyclic octaaza

cobalt(III) complex of sepulcrhate14 L16. This was produced by mixing

[Co(en)3]3+, ammonia and formaldehyde in aqueous solution kept strongly

basic using Li2CO3. The equivalent chemistry was also carried out on the

[Co(sen)]3+ complex (sen=L17) to produce the heptaaza [Co(L18)]3+ 15 and on

10

the [Co(L19)]3+ complex 16 to produce heptaaza [Co(L20)]3+ . In the latter two

cases, one side of the molecule is already 'capped' with a saturated carbon

framework, making them unavailable for facile aminol or imine formation.

Condensation of the formaldehyde molecules with three primary amines on an

octahedral face produces three imines (or aminols) disposed in such a manner

that they can readily undergo reaction with an ammonia molecule which

effectively 'caps' the face by formation of three -NH-CH2-N< links. Although

this is presumably a stepwise and not concerted process, in the presence of

excess formaldehyde and ammonia the reaction proceeds to completion in high

yield under mild conditions. In all of the above complexes, the new aza caps

formed by the condensations do not involve coordination of the 'cap' nitrogen to

the metal ion as the lone pair of the nitrogen is exo to the cavity.

The analogous chemistry to that reported with sen was also achieved with the

N3S3 donor set ligand 'ten' L21 in which the thioether donors play no part. This

produces the macrobicyclic complex of azacapten L2217.

11

Similar chemistry can be promoted around labile metal ions, although little

work involving condensations using formaldehyde and ammonia alone has

been reported. Suh et al.18 have produced an interesting hexaazaalkane L23 by

reacting formaldehyde, ammonia and ethane-1,2-diamine in the presence of

nickel(II). Also produced in this reaction is the hexadentate tripodal heptaaza

ligand L2419,20 which can be seen to be the half-capped analogue of L16

mentioned earlier. However, when heated with excess formaldehyde and

ammonia in dimethyl sulfoxide it readily converts to its square planar co-

product [Ni(L23)]2+, rather than producing the macrobicycle in any significant

yield.

Bernhardt et al.21 have also described the orange complex [Ni(L23)]2+ formed

by reacting [Ni(en)3]2+ (en = ethane-1,2-diamine) with ammonia and

12

formaldehyde solution at room temperature, together with another yellow

square planar product [Ni(L25)]2+. They also found that the equivalent

chemistry can be performed using copper(II) as a template, producing the

additional species [Cu(L26)]2+. The high-field ligands generated in these

reactions promote square-planar geometry for Ni(II) rather than octahedral,

driving the conversion of L24 to L23 reported above, whereas Cu(II) has a

strong inherent preference for square planar geometry. The facility of the

formation of these methylene links to an ammonia ion are shown by the ready

formation of molecules such as L25 and L26.

In place of ammonia, unbound amines can participate in condensation

reactions, although few examples of this type have appeared. One pendant

macrocycle of this type was reported by Suh et al.22. They synthesised the

nickel(II) complex of the ligand L27 by reacting formaldehyde and

methylamine in the presence of the nickel(II) complex of L23. Here a pendant

methyl group is attached to the apical nitrogen in the new 6-membered chelate

ring of the product, with the two protons of the methyl amine replaced in

reaction with two adjacent imines.

13

More recently Suh et al.23 have extended this chamistry by replacing ammonia

with ethanolamine and using Au(III) as a template produce the potentially

sexidentate ligand L28.

In similar reactions, Rosokha et al.24 have reported the condensation of the

nickel(II) complex of 3,7-diazanonan-1,9-diamine with formaldehyde and

aliphatic diamines to produce bis(pentaazamacrocyclic) complexes which

consist of two 14-membered macrocycles joined by an aliphatic bridge between

the non coordinating nitrogen of the new chelate rings L29.

14

1.4 Reaction Involving Formaldehyde and Monofunctional

Methylene Compounds

To date, essentially all reactions in this category have involved the

nitroalkanes [R-CH2-NO2], since these molecules exhibit sufficient acidity in

the methylene adjacent to the nitro group to act, in basic solution, as efficient

nucleophiles for reaction with imines generated from formaldehyde

condensation with coordinated amines. The reaction proceeds as defined in eq.

(7), where coordination of the imine nitrogen to a metal ion during the reaction

is implied.

R-N=CH2 + -CH(R)-NO2 R-NH-CH2-CH(R)-NO2 (7)

The arm generated in this reaction is still an efficient carbon acid, and can

participate in further condensation reactions, at one site if R = alkyl or aryl

group, and at two sites if R= H. Consequences of these addition reactions

where they are internalised in the same molecule are discussed below.

15

1.4.1 Nitromethane

Sargeson and co-workers have expanded their work with cryptates by using

nitromethane as an alternative to ammonia as a nucleophile in the

condensation with formaldehyde. This still allows condensation with three

facially-disposed amines and three formaldehyde molecules, but results in the

apical atom being a carbon with an attached pendant nitro group rather than a

tertiary nitrogen. The pendant nitro group can then be easily reduced to a

primary amine providing a site for additional chemistry, and an additional

potential coordination site.

The reaction between [Co(en)3]3+, formaldehyde and nitromethane in aqueous,

basic solution produced the cobalt(III) complex of the macrobicyclic hexamine

diNOsar L3025 in good yield, analogous to the synthesis of the sepulchrate

([Co(L16)]3+). The ligand is formed in a stepwise process via the condensation

of three formaldehyde units and one nitromethane at each end of the molecule

via formation of imine intermediates. Also produced in low yield (~2%) by the

reaction was the cobalt(III) complex of the half capped analogue NOsen L31 in

which only one end of the molecule is capped, giving some indication of the

probable stepwise nature of the reaction.

Once again, the reaction has been repeated with a large number of analogous

complexes in which one end of the molecule was unavailable for capping,

usually with cobalt(III) acting as a template. These included the capping of the

cobalt complexes of sen L17,26 ten L2527 and taetacn L1928 to produce the

cobalt(III) complexes of L32, L33 and L34. The chemistry has also been

16

described with analogous octahedral complexes of inert metal ions such as

Rh(III), Ir(III) and Pt(IV).29, 30

In an attempt to produce a more rigid cryptand Geue et al.31 substituted

trans-cyclohexane-1,2-diamine for ethane-1,2-diamine in the precursor

cobalt(III) complex. This produced the complex of L34 and the monocapped

analogue [Co(L36)]3+. The reaction was found not to proceed as easily as that

using [Co(en)3]3+ as starting material, and at equivalent temperatures only the

monocapped species was produced in reasonable yield. To produce reasonable

yields of the dicapped species it was necessary to use higher temperatures.

17

In a reversal of the norm, Gainsford et al.32 carried out a synthesis on what

could be conceptualised as a cryptand with the sides removed. They reacted

[Co(tame)2]3+ L37 with formaldehyde and nitromethane to produce the highly

constrained macrotricyclic complex L38 in reasonable yield (ca. 50%). The

ligand has been formed by three formaldehyde and one nitromethane molecule

condensing to form a trigonal cap on one side of the complex, and two

formaldehyde units condensing on the other side of the molecule to form two

four-membered chelate rings fused at a tertiary nitrogen. Two other products

L39 and L40 were found in minor yield (10%). In both, a trigonal cap still

occupies one side of the molecule; however on the other side only one

methylene bridge remains. The other has been replaced by a six member

chelate ring with a pendant nitro group on the ring carbon opposite to the

metal. This has been formed by the condensation of two formaldehyde units

and one nitromethane, and varies in the two molecules in its mode of

attachment. In the former, both methylenes are attached to amines of the

same 'tame' molecule while in the later there is one methylene attached to

each 'tame' molecule.

18

Further work on the same reaction by Geue et al.33 found the additional

product L41, which is essentially similar to the complex L38, but with an

additional fused 4-membered chelate ring.

A variation on this theme was the use of [Co(tach)2]3+ L42 in a condensation

reaction by Geue et al34. The product, L43, shows the formation of a trigonal

cap composed of three formaldehyde units and one nitromethane on one

19

octahedral face, as occurred in the reactions using 'tame'. However, on another

face the capping process has been intercepted to form a stable carbanion

chelate and to methylate the remaining tridentate amine. The maintenance of

this deprotonated form (retained in acid solution) may be due to delocalisation

and the strain engendered in chelating the relatively rigid cage to the metal

ion.

One of the only condensations around a labile metal ion using nitromethane

reacted the copper(II) complex of 4,7-diazadecane-1,10-diamine with

nitromethane and formaldehyde in methanol to produce the 15-membered

macrocycle L44.35 The alcohol pendant is probably the result of an immediate

reaction between a formaldehyde molecule and the acidic proton at the site

formed following condensation of two formaldehyde units and one

nitromethane. Whereas all the acidic protons of nitromethane are replaced by

reaction on a trigonal face, reaction in a plane with only two amines requires

only two acidic protons to be employed. In the presence of excess formaldehyde,

the -NH-CH2-CH(NO2)-CH2-NH- bridge readily undergoes additional reactions

at the central carbon. For reactions directed in one plane, i.e. around metal

20

ions preferring square planar geometry, only molecules of the type R-CH2-NO2

are required, as addressed below.

1.4.2 Nitroethane

Reactions similar to those above around labile metal ions such as Cu(II) and

Ni(II) do not yield macrobicycles in any measurable yields. However, it has

been found that facile high yielding syntheses of monomacrocycles are

possible. One of the most carefully examined series of reactions has involved

use of formaldehyde and nitroethane. This latter reagent may condense with

two formaldehyde residues and two cis-disposed amines (via the imine

intermediate) in the presence of base to form a six membered chelate ring with

pendant nitro and methyl groups attached to the central carbon atom opposite

the metal in the chelate ring. This cyclization has been used extensively to

synthesise a range of cyclic and acyclic polyamine ligands, especially using

copper(II) as a template. As with nitromethane, the nitro pendant group may

be easily reduced to an amine which, after the replacement of the original

21

square planar template with a metal ion preferring octahedral geometry, may

be used as a ligating arm to encapsulate the metal ion.

A good example of a ligand capable of this, and also one of the first examples of

this type of ligand produced using nitroethane and formaldehyde was

synthesised around a [Cu(en)2]2+ template.36 The macrocyclic ligands L45 and

L46 and the half capped analogue L47 were produced around the copper(II)

template in the presence of base (triethylamine) in methanol. It was also found

that, by carefully controlling the ratio of reagents and the rate of addition of

formaldehyde, it is possible to produce almost exclusively either the

macrocyclic or half-capped species. The species L45 and L47 can be viewed as

'planar' analogues of L30 and L31 which form on the trigonal face of an inert

octahedral complex, the former being directed by the preferred square planar

geometry of the metal ion and the choice of carbon acid.

In the synthesis of the macrocycle two generic isomers are possible, with the

methyl (and nitro) groups in syn L46 or anti L45 dispositions. The majority

(~80%) has anti geometry. As the hydrochloride salt of the free hexamine

ligand obtained following zinc reduction, this anti form is markedly less

soluble, allowing the two to be separated. The stereoselectivity towards the

anti form is thought to be due in part to a weak axial interaction between the

22

copper ion and one of the nitro groups during formation directing the

chemistry preferentially toward the anti isomer.

The equivalent chemistry has been extended to the copper(II) complexes of a

range of tetraazaalkanes to produce 13-,14-,15- and 16- membered macrocycles

and a strapped 15-membered macrocycle (L4837, L4938,L5039, L5140, L5241).

Interestingly, unlike the ether analogues produced using formaldehyde alone

(L12-L15) the 13- and 16- membered macrocycles form easily in good yield,

showing the efficacy of the synthetic procedure. There is little variation in the

yield for the reaction with varying ring size; however, the synthesis of the 14-

and 15-membered species proceeds noticeably more easily than the other

reactions, suggesting that the thermodynamic stability of the product is of

importance to the success of these reactions. Following a route similar to that

used to produce L52, Hancock and co-workers have produced analogous

reinforced macrocycles with different sized cyclic systems42,43(L53, L54).

23

The procedure was also applied to the (RR,RR)- [or (SS,SS)-] and (RR,SS)-

[Cu(chxn)2]2+ complex (chxn = trans-cyclohexane-1,2-diamine) to produce the

cyclam analogue L55 and the half capped species L56. Interestingly, the

various chiral forms of the precursor complex tend to produce different ratios

of the cyclic and acyclic forms. This has been attributed to spatial constraints

enforced on the amines by the various isomers44, since the central carbon atom

of the new six-membered chelate ring is a pseudoasymmetric centre in the

(RR,SS)-isomer, and disposition of the coordinated chxn residues varies with

the chirality in the residues and of this centre. In a manner analogous to L45,

the cis isomer of L55 with nitro groups disposed on the same side of the

macrocyclic plane has also been isolated in proportions similar to that found

for the L45, L46 system.45 The closely related asymmetric complex L57 has

also been produced by beginning with a coper complex containing one ethane

diamine unit and one 2,3 diaminocyclohexane unit.45

The same procedure has been applied to a number of copper(II) complexes

with other donors in addition to a pair of cis-disposed amines to produce a

24

range of corresponding cyclic and acyclic species (L5846, L5947, L6048,

L6149,L6245). It is interesting to note that L61 is the result of a two step

synthesis in which the immediate precursor is the acyclic complex of L23,

formed first by the reaction of formaldehyde and ammonia on [Cu(en)2]2+ , as

described above.

While the majority of the complexes produced using formaldehyde and labile

metal templates have been mononuclear, some binuclear species templated by

copper(II), nickel(II) and palladium(II) metal ions have been produced. The

first synthesis was performed by condensation of bis(1,5-diaminopentan-3-

ol)dicopper(II) ion L6350 with formaldehyde and nitromethane in basic

methanol to produce the binucleating macromonocyclic complex L64. A similar

reaction was carried out on the bis(1,5-diaminopentan-3-thilato)dinickel(II) ion

L65 which is analogous to L63, but with the O- donors replaced by sulfur

atoms. The result was a similar binuclear complex L6651, except that the

former reaction yielded L64 as the anti isomer, whereas the latter reaction

yielded L66 as the syn isomer. Although it was not clear that the isomers were

formed stereospecifically, they were formed dominantly. Why the specificity is

directed differently in each case is not clear, although the product L66 adopts a

25

'butterfly' geometry with both Ni(II) centres 'hinged' through the S atoms, and

the -NO2 groups are weakly coordinated in axial sites. Distortion towards this

symmetry in the precusor may drive the outcome. In contrast, L64 has both

Cu(II) centres and all donors in a common plane. The nickel(II) chemistry

above is reproduced with palladium(II) as the templating metal.52

Recently, palladium(II) has also been shown to be an efficient template for

some of the related mononuclear reactions described above.53 Pd(II) forms

much stronger complexes than the labile Cu(II) ion that do not tend to

decompose under the basic conditions employed in these reactions. This has

allowed the synthesis of a number of ‘swollen’ macrocycles (16 and 18

membered macrocycles) which have proven to be difficult to produce using the

more conventional copper or nickel templates (L67, L68, L69).

26

Interestingly the stereo selectivity of the reactions seems greater with the

Pd(II) template with L67 forming almost solely as the trans isomer depicted.

This is in contrast with the same ligand (L45) produced around a Cu(II)

template where appreciable amounts of the cis form (L46) were formed. For

the 16 and 18 membered macrocycles (L68 and L69) there is an apparent

reversal with only the cis forms and isolated. The chemistry has also been

extended to Au(III) and Re(III) ions. The gold template forms an analogue of

L46 and the half capped L47, however with both carrying a stable negative

charge on one of the secondary amines.54 When the equivalent chemistry is

attempted around rhenium the metal is ejected from the complex and the

heterocycle 1,4-bis(2’-nitropropanyl)-6-methyl-6-nitro-1,4-diazacycloheptane

(L70) is formed.54

27

The copper(II) complexes of the potentially binucleating ligands L71 and L7255

were treated with formaldehyde and nitroethane in basic aqueous media to

produce the macrobicyclic species L73 and L74. In the case of L71, the

monocapped macrocycle L75 was also isolated. The substituted spiro-bicyclam

ligands L73 and L74 require the two macrocyclic planes to lie approximately

90o to each other, as the central spiro carbon of the molecules is a tetrahedral

centre, the ligands being fully saturated.

28

1.4.3 Other Nitroalkanes

Reaction of a range of nitro substituted molecules of form R-CH2-NO2 (where

R is an alkyl or substituted alkyl group), formaldehyde and base with the

copper(II) and nickel(II) complexes of 4,7-diazadecane-1,10-diamine and 3,6-

diazaoctane-1,8-diamine respectively produced a range of 13- and 14-

membered macrocyclic products (L76 and L77) with one nitro pendant and the

R group remaining as the other pendant. This chemistry is largely unaffected

by choice of R group.

Evidence of this is also given by replacing nitroethane with nitroproane in the

reaction with bisethanediamine copper(II) to produce the analogues of L45 and

L46, L7856 and L79.57 The inclusion of a carbon acid which produces a longer

alkyl pendant has little effect on the ratios of the two isomers produced in the

reaction.

29

1.4.4 Other Monofunctional Reagents

Considering the efficacy of formaldehyde as a reagent in concert with

ammonia, amines and strong carbon acids such as nitroalkanes, it is hardly

surprising that a deal of attention has been paid to possible alternate reagents

that may be used in concert with formaldehyde to produce new species.

Reactions involving formaldehyde and other reagents in organic chemistry are

extensive; however, metal template reactions which involved species similar to

those discussed above are more limited.

Attention has been focussed recently on modifying the caps of the cryptand

species. This is done either to change the actual physical properties of the cage

or to allow further organic chemistry to be carried out at the cap (e.g. to attach

it to larger molecules). Many of the reaction do not involve chemistry of the

type under discussion, but there are some relevant exceptions. One such

reaction has been reported by Daszkiewicz et al.58. They have shown that a

cobalt(III) complex of the pyruvate imine L80 is an effective nucleophile in

condensing with formaldehyde and [Co(sen)]3+, resulting in a functionalised

sarcophagine cage L81. Here, the methyl group of the pyrvate imine is

sufficiently acidic to act as a tribasic carbon acid, analogous to the chemistry

described for nitromethane earlier. The product L81 may readily be reduced

30

and then reoxidised to form a pendant amino acid L82 from which a range of

organic chemistry may proceed.

Another method of adding functionality to the cap of cryptands has been

described by Hohn et al.59. They report the synthesis of cryptand-like

molecules from the imine derivatives of [Co(sen)]3+ (formed through the action

of formaldehyde on the complex) with mixed aldehydes. Reacting the triimine

derivative of sen L17 with ethanal result in a cryptand with a methanal

pendant on the apical carbon of the new cap L83. If one of the imine groups is

removed to leave the diimine derivative of sen as the precursor complex this

may be reacted with a functionalised ethanal ( RCH2CHO, R= H, Me, Ph). An

intermediate such as L84 results from the reaction of the two acidic protons of

the methylene and subsequent intramolecular condensation produces a

monoimine with a standard size 'cap'. When reduction with NaBH4 follows,

this results in the R-group being the apical pendant to the new cap, as in the

cryptand L85. Throughout this chemistry the cobalt ion remains trapped in the

macrobicycle.

31

Another variation based on [Co(L17)]3+ is to react it with formaldehyde and

either AsH360 or PH3

61 in the presence of base. This results in a capped

macrobicycle with either an arsenic L86 or a phosphorous L87 atom occupying

the apex of the new cap in a form similar to the reaction between [Co(L17)]3+,

formaldehyde and ammonia. The products L86 and L87 analogues of L18, but

the reactions are performed under more stringently controlled conditions.

32

1.5 Reactions Involving Formaldehyde and Bifunctional

Methylene Compounds

The acidity of methylene compounds can be enhanced by having two

electron-withdrawing groups bound directly to the active methylene, i.e.

X-CH2-Y molecules. Examples of molecules in this category include

diethyl malonate (EtOOC-CH2-COOEt) and related molecules such as NC-

CH2-COOEt and EtOOC-CH2-NO2. Clearly, the two acidic protons of the

methylene can be employed in condensation reactions, leading to

molecules with two pendants carrying additional functionality, although

there is the prospect of these being involved in additional or alternate

reactions in some cases, as addressed below.

1.5.1 Diethyl Malonate

Diethyl malonate exhibits sufficent acidity , as a result of the two ester

groups, to function as a relatively strong carbon acid. The two methylene

protons can participate in condenstion reactions with two formaldehyde

molecules and two adjacent coordinated amines in the presence of base,

generating a new six membered chelate ring with a -NH-CH2-C(COOEt)2-

CH2-NH- bridge. This is essentially the same behaviour exhibited by

nitroethane. The resultant gem diester is susceptible to both ester

hydrolysis and decarboxylation, leading eventually to a free acid pendant,

possibly by the route in eq. (8).

33

Where three coordinated amines are available on an octahedral face, the

ester groups in diethyl malonate can be involved in a condensation

reaction in addition to the two acidic methylene protons. Following

condensation with two primary coordinated amines and two formaldehyde

molecules, an intermediate with an ester group and a coordinated primary

amine adjacent is formed (eq. 9, centre) which in basic conditions can

condense further to produce a coordinate amide and complete the 'cap'.

This latter reaction has not been observed as an alternative path in eq. (8),

implying that reaction as a carbon acid is more facile.

34

Reactions with labile metal ions such as copper(II) or nickel(II), where

square planarity is preferred, have been examined in some detail. The

higher pKa of the acidic protons in diethyl malonate in comparison to those

of nitroethane means that reactions involving these reagents do not tend

to proceed as easily as those using nitroethane under identical conditions.

Depending upon the conditions of the reaction, the diester may hydrolyse

and/or decarboxylate to form a mixture of acid ester and monoester

products (eq. 8).

The tetramethyl ester cyclam analogue L88 has been prepared using

dimethylmalonate and formaldehyde (diethylmalonate is equally effective)

in a procedure analogous to that used to produce L4562. Both copper(II)

and nickel(II) templates have been used, as well as the alternative carbon

acid diethylmalonate (which produces the tetraethyl ester). In all

instances, however, the yield tends to be low (ca. 10%), with base induced

decomposition of the precursor complex taking place in competition under

the conditions employed. The analogous procedure has also been carried

out on optically active [Cu(chxn)2]2+ using diethylmalonate to produce the

tetraester L89.45

However, the half capped analogue produced by the controlled addition of

diethylmalonate and formaldehyde to a basic methanolic suspension of

bis(ethane-1,2-diamine)copper(II) perchlorate proceeds readily in good

yield (at least 40%)63. The final product L90 has one acid pendant and one

35

ethyl ester pendant group as opposed to the retention of all the ester

groups in the synthesis of the macrocyclic analogue. It is readily

decarboxylated to the monoacid L92. Apart from reactions involving

ethane-1,2-diamine, the basic synthetic method has been applied to a large

variety of polyamines to produce a number of cyclic and acyclic species, all

with either acid or ester pendant groups (L92, L93, L9464, L9565, L9666,

L9745).

Another variation is the use of diethylmalonate and nitroethane

condensation in the same molecule. In a multiple step process commencing

with [Cu(en)2]2+, capping using diethylmalonate was employed to produce

L90. This species is then stabilised by conversion to the methyl ester

36

derivative of L91 by hydrolysis in basic methanol. The other pair of

primary amines of the intermediate complex molecule is then capped

using nitroethane, the methyl ester being hydrolyzed in the process. The

resulting complex L109AA is a cyclam derivative with both nitro and

carboxylate pendant group which, after reduction to convert the nitro to

an amine and to remove the metal, is capable of encapsulating a wide

range of octahedral metal ions as a sexidentate ligand L99.67 Also

produced from this reaction is the analogous cis isomer in proportions

similar to that found for the L45, L46.

Likewise, L88 can be converted by hydrolysis and decarboxylation

reactions to the potentially sexidentate ligand L100 (L89 can also undergo

the analogous process to produce diacid species). Although no cis/trans

isomerisation is possible for the tetraester, the hydrolysis and

37

decarboxylation reactions produces both isomers, with a strong preference

for the trans form (4:1, as is discussed in later chapters). Since L45 can be

readily reduced to the hexamine L101, the three ligands L99L101 (and

there cis analogues) represent a series of potentially sexidentate

macrocyclic tetraamines with two pendant donors, where the pendant

donors step from two amines, to an amine and a carboxylate, to two

carboxylates. The coordination chemistry and synthesis L99 and L100 are

discussed in detail in Chapters 4 and 5 and compared with the well

characterised complexes of L101.

Reaction of inert hexaamine cobalt(III) complexes with formaldehyde and

diethylmalonate have been described only recently. Reaction of Co(en)33+

generates essentially only the mono-capped molecule L102, capping at the

other face of primary amines not being observed. However, commencing

with Co(L17)3+ or Co(L24)3+ yields macrobicyclic amides L103 and L104

respectively. The pendant ester group can be converted via standard

organic chemistry reactions to the acid (via base hydrolysis) and the acid

chloride (via reaction with thionyl chloride), the later then capable of facile

condensation reaction with amines. The half-capped cobalt(III) complex of

L102 can be converted by reaction with formaldehyde and nitroethane to

L105. The coordinated amide is extraordinarily robust, being unaffected

by these reactions, but is also not conveniently reduced.

38

1.5.2 Other Reagents

There are a number of other conveniently available bifunctional

methylene compounds which are sufficiently strong carbon acids to permit

participation in reactions analogous to those described above. The

'composite' molecule EtOOC-CH2-NO2 is a clear candidate, but few

reactions with the molecule have been reported to date, beyond some

evidence for the formation of L10668 in a condensation reaction with

formaldehyde and a tetraamine copper(II) ion.

39

Bifunctional methylene compounds incorporating nitriles are candidates

for condensations, and some examination of NC-CH2-COOEt and NC-CH2-

CN has been performed. The latter appears to yield L107 in condensation

reactions analogous to L45, but the stability of the pendant molecule is not

apparently great and the product has not been definitely confirmed. The

former reagent undergoes an analogous reaction to form L108.45 The

former reagent also undergoes an analogous reaction to diethyl malonate

with Co(L17)3+ to produce L109. In principle, provided acidity is great

enough, reactions should proceed essentially as defined for diethyl

malonate, although the possible diversion of the reaction along alternative

paths is not well understood, and offers avenues for further examination.

40

1.6 Conclusions

The synthesis of macrocycles, particularly saturated macromonocycles and

macrobicycles incorporating mainly nitrogen donors, is often facile when

inert or highly stable labile complexes with appropriately disposed

coordinated primary amines are reacted with formaldehyde and ammonia

or functionalised methylene compounds which are efficient carbon acids.

Around inert octahedral metal ions, saturated and usually macrobicylcic

molecules are readily accessible. Around more labile metal ions, such as

copper(II), nickel(II) and palladium(II), there is a tendency to form usually

saturated macromonocyclic molecules, directed by a preference for four-

coordination and square planarity with strong-field ligands in those cases.

In many cases, the resultant ligand can be removed from the metal ion in

an intact or slightly amended form. For example, copper(II) complexes of

macrocycles produced from reactions described above may often be

obtained simply by a zinc/acid reduction reaction with following

chromatography on Dowex resin to isolate the metal-free pendant-arm

macrocycle in high yield.69 Where inert complexes such as cobalt(III)

complexes of macrobicycles are produced, the metal-free macrobicyclic

ligand may be obtained by reaction under an oxygen-free atmosphere of

the reduced cobalt(II) complex either with concentrated hydrobromic acid

41

or with excess cyanide ion, which remove the cobalt(II) as CoBr42- or

Co(CN)42- respectively.70

The capacity of the pendant-arm macromonocycles and macrobicycles to

bind metal ions strongly and in some cases selectively, and the unusual

physical properties often shown by complexes of these isolated ligands,

provides a wealth of additional chemistry which is still under active

investigation. In the following chapters, the synthesis and coordination

chemistry of a number of the macromonocycles bearing pendant

carboxylates discussed is described and compared with the already

extensive information concerning their amine pendant analogues.

1.7 References

1. R. D. Hancock and A. E. Martell, Chem. Rev., 89 (1989) 1875 and references

therein.

2. G. A. Melson (ed), 'Coordination Chemistry of Macrocyclic Complexes', Plenum,

New York (1982); M. deS. Healy and A. J. Rest, Adv. Inorg. Chem. Radiochem., 21

(1978) 1

3. I.I. Creaser, J. M. Harrowfield, A.J.Herlt, A. M. Sargeson, J. Springborg, R. J.

Geue, M. R. Snow, J. Am. Chem. Soc., 99 (1977) 3181

4. N. W. Alcock and P. A. Tasker, J. Chem. Soc., Chem. Commun., (1972) 1239

(1972)

5. S. M. Peng, G. C. Gordon and V. L. Goedken, Inorg. Chem., 17 (1978) 119

42

6. R. J. Geue, M. G. McCarthy, A. M. Sargeson, B. W. Skelton and A. H. White,

Inorg. Chem., 24 (1985) 1607

7. G. A. Lawrance, T. M. Manning, B. W. Skelton, J. Chem. Soc., Chem. Commun.,

(1988) 1344

8. M. P. Suh and S-G. Kang, Inorg. Chem., 27 (1988) 2544

9. M. P. Suh, W. Shin, S-G. Kang, M. S. Lah and T. M. Chung, Inorg. Chem., 28

(1989) 1602

10. M. P. Suh, S-G. Kang, V. L. Geodken and S-H. Park, Inorg. Chem., 30 (1991) 365

11. J. R. Brush, R. J. Magee, M. J. O'Connor, S.B.Teo, R.J. Geue and M. R. Snow, J.

Am. Chem. Soc., 95 (1973) 2034

12. R. J. Geue, M. R. Snow, J. Springborg, A. J. Herlt, A. M. Sargeson and D. Taylor,

J. Chem. Soc., Chem. Commun., (1976) 285

13. G. A. Lawrance, M. Maeder, M. A. O'Leary, B. W. Skelton and A. H. White, Aust.

J. Chem., 44 (1991) 1227

14. I.I. Creaser, J. M. Harrowfield, A.J.Herlt, A. M. Sargeson, J. Springborg, R, J,

Geue, M. R. Snow, J. Am. Chem. Soc., 99 (1977) 3181

15. I.I. Creaser, J. M. Harrowfield, A.J.Herlt, A. M. Sargeson, J. Springborg, R. J.

Geue, M. R. Snow, J. Am. Chem. Soc., 104 (1982) 6016

16. A. Hammershoi and A. M. Sargeson, Inorg. Chem., 22 (1983) 3554

17. L. R. Gahan, T. W. Hambley, A. M. Sargeson and M. R. Snow, Inorg. Chem., 21

(1982) 2699

18. M. P. Suh, W. Shin, H. Kim and C. H. Koo, Inorg. Chem., 26 (1987) 1846

19. M. P. Suh, W. Shin, D. Kim and S. Kim, Inorg. Chem., 23 (1984) 618

20. M. P. Suh and D. Kim, Inorg. Chem., 24 (1985) 3712

43

21

. P. V. Bernhardt, L. S. Curtis, N. F. Curtis, G.A. Lawrance, B. W. Skelton and A.

H. White, Aust. J. Chem., 42 (1989) 797

22. M. P. Suh, W. Shin, H. Kim and C. H. Koo, Inorg. Chem., 26 (1985) 30

23 M. P. Suh, I. S. Kim, B. Y. Shim, D. Hong and T. Yoon, Inorg. Chem., 35 (1996)

3595

24. S. V. Rosokha and Y. D. Lampeka, J. Chem. Soc., Chem. Commun., (1991) 1077

25. R. J. Geue, T. W. Hambley, J. M. Harrowfield, A. M. Sargeson and M. R. Snow, J.

Am. Chem. Soc., 106 (1984) 5478

26. A. M. Sargeson, Chem. Brit., 15 (1979) 23

27. L. R. Gahan, T. W. Hambley, A. M. Sargeson and M. R. Snow, Inorg. Chem., 21

(1982) 2699

28. A. Hammershoi and A. M. Sargeson, Inorg. Chem., 22 (1983) 3554

29. H. A. Boucher, G. A. Lawrance, P. A. Lay, A. M. Sargeson, A. M. Bond, D. F.

Sanster and J. C. Sullivan, J. Am. Chem. Soc., 105 (1983) 4652

30. P. A. Lay, doctoral dissertation, The Australian National University (1981)

31. R. J. Geue, M. G. McCarthy and A. M. Sargeson, J. Am. Chem. Soc., 106 (1984)

8282

32. G. J. Gainsford, R. J. Geue and A. M. Sargeson, J. Chem. Soc., Chem. Commun.,

(1982) 223

33. R. J. Geue, M. G. McCarthy, A. M. Sargeson, B. W. Skelton and A. H. White,

Inorg. Chem., 24 (1985) 1607

34. R. J. Geue, M. G. McCarthy, A. M. Sargeson, E. Horn and M. R. Snow, J. Chem.

Soc. Chem. Comm., (1986) 848

35. G. A. Lawrance, M. Maeder, M. O'Leary and F. Woon, Polyhedron, 9 (1990) 2227

44

36

. P. Comba, N. F. Curtis, G. A. Lawrance, A. M. Sargeson, B. W. Skelton and A. H.

White, Inorg. Chem., 25 (1986) 4260

37. P. Comba, N. F. Curtis, G. A. Lawrance, M. O'Leary, B. W. Skelton and A. H.

White, J. Chem. Soc., Dalton Trans., (1988) 497

38. G. A. Lawrance, M. Rossignoli, B. W. Skelton and A. H. White, Aust. J. Chem., 40

(1987) 1441

39. G. A. Lawrance and M. O'Leary, Polyhedron, 6 (1987) 1291

40. P. Comba, N. F. Curtis, G. A. Lawrance, M. O'Leary, B. W. Skelton and A. H.

White, J. Chem. Soc., Dalton Trans., (1988) 2145

41. L. M. Engelhardt, G. A. Lawrance, T. M. Manning and A. H. White, Aust. J.

Chem., 42 (1989) 1859

42. G. Pattrick, M. P. Ngwenya, S. M. Dobson and R. D. Hancock, J. Chem. Soc.,

Dalton Trans., (1991) 1295

43. G. Pattrick and R. D. Hancock, Inorg Chem, in press

44. P. V. Bernhardt, P. Comba, B. Elliot, G. A. Lawrance, M. Maeder, M. A. O'Leary,

G. Wei and E. N. Wilkes, Aust. J. Chem., 47 (1994) 1171

45 G. Wei, Doctoral Dissertation, University of Newcastle, 1997

46. P. Comba, T. W. Hambley and G. A. Lawrance, Helv. Chim. Acta, 68 (1985) 3232

47. G. A. Lawrance, T. M. Manning and M. O'Leary, Polyhedron, 7 (1988) 1263

48. P. V. Bernhardt, L. S. Curtis, N. F. Curtis, G. A. Lawrance, B. W. Skelton and A.

H. White, Aust. J. Chem., 42 (1989) 797

49. P. Comba, G. A. Lawrance, M. Rossignoli, B. W. Skelton and A. H. White, Aust.

J. Chem., 41 (1988) 773

50. P. Comba, G. A. Lawrance, T. M. Manning, A. Markiewicz, K. S. Murray, M. A.

O'Leary, B. W. Skelton and A. H. White, Aust. J. Chem., 43 (1990) 69

45

51

. G. A. Lawrance, M. Maeder, T. M. Manning and M. A. O'Leary, B. W. Skelton

and A. H. White, J. Chem. Soc., Dalton Trans., (1990) 2491

52. G. A. Lawrance, M. Maeder, T. M. Manning, M. O'Leary, B. W. Skelton and A. H.

White, J. Chem. Soc., Dalton Trans., (1990) 2491

53. M. Rossignoli, C. C. Allen, T. W. Hambley, G. A. Lawrance and M. Maeder,

Inorg. Chem. 35 (1996) 4961

54 Monica Rossignoli, Doctoral Disertation, University of Newcastle, 1997.

55. P. V. Bernhardt, P. Comba, L. R. Gahan and G. A. Lawrance, Aust. J. Chem., 43

(1990) 2035

56 N.F. Curtis, G. J. Gainsford, A. Siriwardena and D. C. Weatherburn, Aust J. Chem.,

46 (1993) 755

57 P. V. Bernhardt, K. A. Byriel, C. H. L. Kennard and P. C. Sharpe, Inorg. Chem., 35

(1996) 2045

58. B. Korybut-Daszkiewicz, R. M. Hartshorn and A. M. Sargeson, J. Chem. Soc.,

Chem. Commun., (1989) 1357

59. A. Hohn, R. J Geue and A.M. Sargeson, J. Chem. Soc., Chem. Commun., (1990)

1473

60. A. Hohn, R. J. Geue, A. M. Sargeson and A. C. Willis, J. Chem. Soc., Chem.

Commun., (1989) 1648

61. A. Hohn, R. J. Geue, A. M. Sargeson and A. C. Willis, J. Chem. Soc., Chem.

Commun., (1989) 1644

62. L. Xin, N. F. Curtis and D. C. Weatherburn, Trans. Met. Chem., 17 (1992) 147

63. P.V. Bernhardt, G. A. Lawrance, B.W. Skelton and A. H. White, Aust. J. Chem.,

43 (1990) 399

46

64

. G. A. Lawrance, B. W. Skelton, A. H. White and E. N. Wilkes, Aust. J. Chem., 44

(1991) 1511

65. G. A. Lawrance and M. O'Leary, Polyhedron, 6 (1987) 1291

66. G. A. Lawrance and E. N. Wilkes, unpublished results

67. T. W. Hambley, G. A. Lawrance, M. Maeder and E. N. Wilkes, J. Chem. Soc.,

Dalton Trans., (1992) 1283

68 G. A. Lawrance and M. A. O'Leary, unpublished results

69. G. A. Lawrance, T. M. Manning, M. Maeder, M. Martinez, M. A. O'Leary, W. C.

Patalinghug, B. W. Skelton and A. H. White, J. Chem. Soc., Dalton Trans., (1992)

1635

70. A. M. Sargeson, Pure Appl. Chem., 56 (1984) 1603; Chem. Aust., (1991) 176